invgamA: Bivariate Archimedean copula based on inverse gamma LT

In YafeiXu/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description

Bivariate Archimedean copula based on inverse gamma Laplace transform

Usage

pinvgamA(u,v,cpar)
dinvgamA(u,v,cpar)
pcondinvgamA(v,u,cpar)  # C_{2|1}(v|u;cpar)
rminvgamA(n,d,cpar)
logdinvgamA(u,v,cpar,pgrid=0)
invgamA.cpar2tau(cpar)
invgamA.tau2cpar(tau)

Arguments

u	value in interval 0,1; could be a vector
v	value in interval 0,1; could be a vector
cpar	parameter: could be scalar or vector (positive-valued)
n	sample size for ripsA, positive integer
d	dimension
pgrid	grid of values in (0,1) to use for monotone interpolation; see code for the default vector when pgrid is input as 0
tau	Kendall value in (0,1)

Value

cdf, pdf, conditional cdf, conditional quantile value(s) for pinvgamA, dinvgamA, pcondinvgamA, qcondinvgamA respectively;

log density for logdinvgamA (use for maximum likelihood);

random d-vectors for rminvgamA;

Kendall's tau for invgamA.cpar2tau;

copula parameter for invgamA.tau2cpar

References

Joe H and Hua L (2010). Tail order and intermediate tail dependence of multivariate copulas. Journal of Multivariate Analysis, v 102, 1454–1471

Examples

u=seq(.1,.6,.1)
v=seq(.4,.9,.1)
cpar=.5
pp=pcondinvgamA(v,u,cpar)
print(pp)
tau=invgamA.cpar2tau(cpar)
print(tau)
set.seed(123)
udata=rminvgamA(500,d=2,cpar=2) # tau=0.5
print(taucor(udata[,1],udata[,2]))
print(semicor(udata,inscore=FALSE))
ml=nlm(bivcopnllk,p=1.5,hessian=TRUE,print.level=1,
  udat=udata,logdcop=logdinvgamA,LB=.0001,UB=10)
## Not run: 
# contour of density with N(0,1) margins
zvec=seq(-3,3,.2)
contourBivCop(2,zvec,dinvgamA)
## End(Not run)

YafeiXu/CopulaModel documentation built on May 9, 2019, 11:07 p.m.